You are here

Article Text

Article menu

Download PDFPDF

Research report
Social determinants of mental health: a Finnish nationwide follow-up study on mental disorders
  1. Reija Paananen1,
  2. Tiina Ristikari1,
  3. Marko Merikukka1,
  4. Mika Gissler2,3
  1. 1Department of Children, Young People and Families, National Institute for Health and Welfare, Oulu, Finland
  2. 2Information Department, National Institute for Health and Welfare, Helsinki, Finland
  3. 3NHV Nordic School of Public Health, Gothenburg, Sweden
  1. Correspondence to Dr Reija Paananen, Department of Children, Young People and Families, Finnish National Institute for Health and Welfare, P.O. Box 310, Oulu 90101, Finland; reija.paananen{at}thl.fi

Abstract

Background Most mental disorders start in childhood and adolescence. Risk factors are prenatal and perinatal, genetic as well as environmental and family related. Research evidence is, however, insufficient to explain the life-course development of mental disorders. This study aims to provide evidence on factors affecting mental health in childhood and adolescence.

Data and methods The 1987 Finnish Birth Cohort covers all children born in Finland in 1987 (N=59 476) who were followed up until the age of 21 years. The study covers detailed health, social welfare and sociodemographic data of the cohort members and their parents from Finnish registers.

Results Altogether, 7578 (12.7%) cohort members had had a diagnosed mental disorder. Several prenatal, perinatal and family-related risk factors for mental disorders were found, with sex differences. The main risk factors for mental disorders were having a young mother (OR 1.30 (1.16 to 1.47)), parents’ divorce (OR 1.33 (1.26 to1.41)), death of a parent (OR 1.27 (1.16 to 1.38)), parents’ short education (OR 1.23(1.09 to 1.38)), childhood family receiving social assistance (OR 1.61 (1.52 to 1.71)) or having a parent treated at specialised psychiatric care (OR 1.47 (1.39 to 1.55)). Perinatal problem (OR 1.11 (1.01 to 1.22)) and prenatal smoking (OR 1.09 (1.02 to 1.16)) were risk factors for mental disorders, even after controlling for background factors. Elevated risk was seen if the cohort member had only basic education (OR 3.37 (3.14 to 3.62)) or had received social assistance (OR 2.45 (2.30 to 2.60)).

Conclusions Mental disorders had many social risk factors which are interlinked. Although family difficulties increased the risk for mental disorders, they were clearly determined by the cohort member's low education and financial hardship. This study provides evidence for comprehensive preventative and supporting efforts. Families with social adversities and with parental mental health problems should be supported to secure children's development.

  • Mental Health
  • Social and Life-Course Epidemiology
  • Public Health

https://doi.org/10.1136/jech-2013-202768

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Mental health in childhood and adolescence is an increasing public health concern since most adulthood mental disorders begin in childhood and adolescence.1 Based on epidemiological studies, it is estimated that at least one of every four to five young people suffers from at least one mental disorder in the general population.2 Mental disorders contribute to significant suffering, functional impairment, exposure to discrimination and enhanced risk of premature death.3

    There is ample evidence showing how mental disorders have roots in childhood.4 The aetiology is complex and multifactorial, involving biological, social and behavioural mechanisms operating across the life course. Previous research has identified various prenatal and perinatal, genetic as well as environmental and family-related risk factors for mental disorders,5 ,6 such as genetic vulnerability, stressful life events, low social cohesion, physical health problems, factors related to the family, like low education, lack of employment and factors related to family function.7–12 Research evidence is, however, insufficient to explain the life-course development of mental disorders, and knowledge is scarce about the joint effect or the synergism of the various risk factors.

    This longitudinal study aims to provide evidence on factors affecting mental health in childhood and in adolescence. The aim of this study was to assess the risk for mental disorders by analysing the effects of prenatal and perinatal factors, family stability, education and financial problems by sex. We also investigated how the intergenerational transmission of education and financial problems is associated with mental disorders. This knowledge of the potentially preventable factors is needed for comprehensive preventative and supporting efforts.

    Materials and methods

    The 1987 Finnish Birth Cohort

    The 1987 Finnish Birth Cohort (1987 FBC) is a longitudinal nationwide follow-up data study including a complete census of all infants born in a single year and subsequently followed over time with detailed forms of documentation of their own and their parents’ health status and social circumstances from the perinatal period into early adulthood. The data were gathered from Finnish national registers, which are shown to be of high quality and appropriate for research purposes.13 Altogether, 60 069 children, including all live births and stillbirths of infants weighing more than 500 g, or having a gestational age of 22 weeks or more, born in Finland in 1987 and registered in the Medical Birth Register (MBR) were included in the 1987 FBC follow-up study covering the years 1987–2008.14 Because of the identification number given by the Central Population Register to every newborn and permanent migrant to Finland was incomplete, missing, incorrect or changed, 73 (0.1%) children were untraceable from the national registers. The remaining number of live-born children included in the follow-up study until the end of the year 2008 was 59 669. The children surviving the perinatal period were included in the study (n=59 476). During the study period, 497 (0.8%) participants died and 662 (1.1%) had emigrated permanently by 31 December 2008. These cohort members were included in the study until death or emigration, respectively. At the end of the follow-up, 58 320 cohort members (98.1%) were still alive and living in Finland.

    The study obtained ethics approval of the National Institute for Health and Welfare (Ethical committee §28/2009), and permission to use the register data was obtained from all register keeping organisations.

    Data on psychiatric care and psychiatric diagnoses

    The Finnish Hospital Discharge Register (HDR), kept by THL (National Institute for Health and Welfare), includes all inpatient care episodes from all Finnish hospitals since 1969 and all specialised level outpatient visits in public hospitals since 1998. HDR has been found to be a valid and reliable tool for epidemiological research.15 Data on the cohort member's psychiatric diagnoses were collected from specialised care and reported using the International Classification of Diseases (ICD)-9 in 1987 and 1995 (290–319) and ICD-10 since 1996 (F00-99). Data on parental psychiatric care were collected from HDR, for psychiatric inpatient care between 1 January 1987 and 31 December 2008, and outpatient care in between 1 January 1998 and 31 December 2008. Parental psychiatric care was counted if either parent had at least one outpatient or inpatient visit at psychiatric specialities during follow-up.

    Data on financial hardship

    The recipients of social assistance are registered by THL. Social assistance refers to last-resort financial assistance provided by social services to a household from municipal funds when other sources of income are insufficient to ensure that the basic needs of a person or a family are met. Parental social assistance was used as an indicator of the family financial hardship, and cohort members’ receipt of social assistance as an indicator of their own poor financial situation (classified in analyses yes/no). Parental social assistance was registered for either the biological mother, biological father or for both parents during the follow-up 1987–2008, and for the cohort members 2002–2008.

    Data on family situation

    Data on the cohort members’ biological parents’ mutual marriages and divorces (classified in the analyses as parents married and divorced vs married and not divorced during the follow-up) and information on their death (yes/no) during the follow-up were received from The Finnish Central Population Register. Data on the mother's marital status at the cohort member's birth (classified in the analyses as a single mother including unmarried and not cohabiting at birth vs married or cohabiting) were collected from MBR, kept by THL.

    Prenatal and perinatal data

    Data on mother's age (classified in the analyses as under 20 years at birth vs 20 years or more), maternal smoking during pregnancy (yes/no), mother's previous induced abortions (yes/no) and prenatal and perinatal conditions were obtained from MBR. If the child was born preterm (under 37 gestational weeks), weighed below 2500 g at birth, was diagnosed as small for gestational age, or if the Apgar scores at 1 or 5 min were 1–3, he/she was classified as having a perinatal problem (yes/no).

    Data on parental socioeconomic status and education

    Data on the socioeconomic status (SES) of the cohort members’ biological parents were collected from the Finnish Central Population Register on 10 June 2009, and it included the parents’ most recent occupations, classified as upper white-collar workers, lower white-collar workers, blue-collar workers or ‘others’, including, for example, entrepreneurs, students, housewives and farmers. Information on the highest educational achievements was received from Statistics Finland and classified as high school or higher (12 years or more of education), lowest level tertiary (11 years), lower secondary (10–11 years) or primary (up to 9 years). We considered SES and the educational attainment of each parent. Parent's highest SES and education indicated the highest of either parent (the findings were similar using only father's or mother's SES or education). The cohort members’ own educational achievements were classified as comprehensive school only, upper secondary school or matriculation examination or academic degree.

    Intergenerational variables

    Social assistance receipt and educational attainment were classified as intergenerational variables measuring the intergenerational social mobility. These variables were classified into four categories: (1) neither the parents nor the cohort member had received social assistance (reference), (2) The parents did not receive social assistance, but the cohort member did , (3) The parents had received social assistance, but the cohort member did not and (4) both the parents and the cohort member had received social assistance during the follow-up. Similarly, education was evaluated as an intergenerational variable and classified as follows: (1) both the parents and the cohort member had an educational degree after comprehensive school (reference), (2) the parents had further education, but the cohort member did not, (3) the parents did not have any further education, but the cohort member had and (4) neither the parents nor the cohort member had any further educational degrees.

    Statistical methods

    The register data were combined using the children's and their parents’ personal identification numbers. Logistic regression analyses were used to define the ORs for prenatal and perinatal, family related and cohort member's variables on risk for mental disorders. Six multivariate models were created to evaluate the relationships between the risk factors. The first model evaluated all the prenatal and perinatal variables: mother's young age, single motherhood, smoking during pregnancy, previous induced abortions and perinatal problem. Model 2 evaluated the family-related variables; parental divorce, parental death, parental social assistance receipt, parental psychiatric care, parental education and parental SES. Model 3 combined all the prenatal and perinatal and family-related variables. Model 4 evaluated the effect of intergenerational variables, education and social assistance, and model 5 took into account the cohort member's education and social assistance receipt together with the prenatal and perinatal and family-related variables. The full model included all previously significant variables related to the cohort members and their family as well as the intergenerational variables characterising the influence of social mobility on mental disorders. Second-order interactions were evaluated within the variable groups. The data analysis was performed using PASW Statistics, V.18.

    Results

    Based on the information from the Finnish HDR over the 21-year follow-up from 1987 to 2008, 7578 (12.7%) of 59 476 children from the 1987 FBC had had a diagnosed psychiatric disorder (excluding intellectual disabilities F70–79): 14.2% of the girls and 11.4% of the boys. Three of the most prevalent diagnoses were mood (affective) disorders (F30–39), which were given for 7% of the cohort members, neurotic, stress-related and somatoform disorders (F40–49), 5.6% and behavioural and emotional disorders with onset usually occurring in childhood and adolescence (F90–98), 3%.16 These were also the diagnoses most often showing comorbidity. In 1987, when the cohort member was born, 5.2% of the mothers were single parenting (table 1). In terms of smoking, 15% of the cohort members had been exposed to prenatal maternal smoking, and 14.6% of the cohort members’ mothers had had previous induced abortions. Perinatal problems had been experienced by 7.3% of the cohort members. During the follow-up, 29.6% of the parents were divorced and 4047 (6.8%) members of the cohort had lost one of their parents. The recession of the 1990s shows up in the data when we look at the number of parents who had received social assistance, a total of 38%. Of the cohort members, 22.8% had received social assistance during their young adulthood, girls (24.6%) slightly more often than boys (21.1%). Of those, 26.7% also had a psychiatric diagnosis. Almost three of four (71.6%) members of the cohort who had received social assistance had a parent who had also received social assistance. By the age of 21 years, 18.5% of the cohort members had completed only comprehensive school (boys 20.7% vs girls 16.1%). Furthermore, 38.1% of the cohort members with parents who had basic education only had themselves no further educational degree and 28.3% of those with no education also had a psychiatric diagnosis.

    View this table:
    Table 1

    Determinants for mental disorders in the 1987 Finnish Birth Cohort

    Overall, several prenatal and perinatal, and family-related risk factors for mental disorders were found, and the likelihood for a mental disorder also varied according to sex. Young (under 20 years; OR 1.96 (1.76 to 2.20)) or single mothers (OR 1.66 (1.51 to 1.82)) and maternal smoking during pregnancy (OR 1.74 (1.64 to 1.85)) were the strongest prenatal and perinatal determinants for mental disorder in childhood or in adolescence (table 1). Changes in family situations, like divorce (OR 1.68 (1.60 to 1.76)) or parental death (OR 1.82 (1.68 to 1.97)), were significant determinants for mental disorder, as well as parental low socioeconomic position (OR 1.47 (1.38 to 1.57)) and short education (OR 1.91 (1.74 to 2.10)). Parental financial hardship, measured by social assistance received during the follow-up, was a significant risk factor for children's mental disorder (OR 2.16 (2.06 to 2.27)), as well as parental mental disorders, registered as specialised psychiatric care (OR 1.79 (1.70 to 1.89)). Mother's young age, parental social assistance receipt and parental psychiatric care were more prominent determinants for mental disorders for boys than for girls.

    Parental social assistance receipt, together with the cohort member's own received social assistance, showed a significant risk for psychiatric mental disorder (OR 4.63 (4.36 to 4.91)), as well as parental short education together with the cohort member's own short education (OR 4.18 (3.73 to 4.68)). The cohort member's own social assistance receipt and own short education were, however, more prominent determinants than the parental ones. The risk for a mental disorder was associated with the adolescent's own low education (OR 4.75 (4.47 to 5.04)), more so for boys (5.64 (5.13 to 6.21)) than girls (4.71 (4.34 to 5.11)). In addition, adolescents who had received social assistance during the follow-up had significantly more often a diagnosis for a mental disorder than the other members of the cohort (3.86 (3.67 to 4.05)).

    As shown in table 2, model 1, all the prenatal and perinatal factors, mother's young age, single motherhood, smoking during pregnancy, previous induced abortions and perinatal problem stayed significant, but mother's young age and smoking during pregnancy were the strongest determinants for children's mental disorders. Model 2 included the family-related factors, and divorce, parental death, parental social assistance receipt, psychiatric care and short education, and as shown in table 2, model 2, all these variables stayed statistically significant after adjusting (p<0.001, ORs varying from 1.28 to 1.76), with the exception of parental SES. Parental education together with social assistance receipt was the strongest interaction studied. The risk for mental disorders was higher if parents had received social assistance and had low education.

    View this table:
    Table 2

    Determinants for mental disorders in the 1987 Finnish Birth Cohort, multivariate models

    Model 3 combined all the prenatal and perinatal and family-related variables (table 2, model 3). Mother's previous induced abortions and perinatal problem barely lost their significance, other factors staying as significant risk factors, parental social assistance being the most prominent risk factor (OR 1.61 (1.52 to 1.71)). When evaluating social assistance receipt and education as intergenerational variables (table 2, model 4), the strongest risk for children's mental disorders was seen when both the parents and the cohort member had received social assistance (OR 3.12 (2.92 to 3.30)).

    Model 5 took into account the cohort member's education and social assistance receipt together with the prenatal and perinatal and family-related variables (table 2, model 5), and the results show that the variables related to the cohort member were the strongest determinants for mental disorders. However, maternal smoking (OR 1.10 (1.02 to 1.17)) and perinatal problem (OR 1.11 (1.01 to 1.22)) stayed as significant determinants after adjusting. By sex, perinatal problem was only significant for boys (data not shown). Parental psychiatric care was the strongest family-related determinant for children's mental disorder (OR 1.42 (1.34 to 1.50)). Parental social assistance receipt lost significance just barely, but it stayed as a significant risk factor for boys of the cohort (data not shown), while divorce and death of a parent stayed significant for both sexes. In this model, parental short education was no longer a significant risk factor.

    Interestingly, when taken into account, all significant variables related to prenatal and perinatal conditions, family and cohort members, social assistance and education as intergenerational variables; by logistic modelling (table 2 full model), the cohort member's short education was a more significant determinant for mental disorder when the parents were educated (OR 2.80 (2.63 to 2.97)) than when both the parents and the cohort member had only received basic education (OR 2.31 (2.03 to 2.61)). In addition, the risk for mental disorder was higher if both the parents and the cohort member were receivers of social assistance. Parental social assistance receipt was a significant risk factor for mental disorders among boys, but not for girls, regardless of whether the cohort member had received social assistance or not (data not shown). Female sex (OR 1.33 (1.26 to 1.40)) and parental psychiatric care (OR 1.42 (1.34 to 1.50)), maternal smoking (OR 1.09 (1.02 to 1.16)), perinatal problem (OR 1.11 (1.01 to 1.22)), as well as parents’ divorce (OR 1.19 (1.12 to 1.26)) and parental death (OR 1.28 (1.17 to 1.41)) stayed as significant risk factors for mental disorders when adjusted with all prenatal and perinatal, family-related and intergenerational variables.

    Discussion

    This study reveals that 13% of the young adults born in Finland in the year 1987 had had a mental disorder diagnosed in specialised healthcare during the follow-up from childhood until early adulthood. With the longitudinal, nationwide 21-year follow-up, we show that mental disorders, educational difficulties and lack of income are interwoven and that mental disorders in childhood and in adolescence are largely determined by social factors.

    The main family-related risk factors for children's mental disorders were parents’ divorce, death of a parent, financial hardship of the childhood family or having a parent with psychiatric problems, all of which are known risk factors from the previous epidemiological research as well.4 Similar longitudinal research, for example, in the USA has shown how family finances are an important predictor for the development of youth. Poverty impacts children's learning abilities most, and slightly less physical health and behaviour.17 ,18 This study also confirms the previous evidence that parental psychiatric problems have clearly a strong adverse impact on their children's mental health.4 ,19–21 The risk of parental psychiatric care on children's mental disorder diminished by a fifth when controlled by other prenatal and perinatal, family-related or cohort member's factors, but stayed almost constant despite the nature of confounding factors, indicating the independent nature of parental psychiatric problems to their offspring.

    Maternal smoking during pregnancy showed as a significant risk factor for mental disorders, even after controlling for other social and family-related factors. In recent years, research evidence has accumulated, showing the independent risk and potential biological pathways of prenatal smoking on children's and adolescent's mental health problems.22–26 Even if it is difficult to specify the impact of prenatal smoking separately from social determinants, maternal smoking during pregnancy is a preventable prenatal risk factor and should be highlighted.

    Although family difficulties increased children's risk for mental disorders, they were clearly determined by adolescents’ own low education and financial hardship. It is known that mental disorders in young people have a substantial effect on social and economic outcomes that extend into adulthood,27 and that mental disorders that begin in childhood have an impact on school performance and academic achievement.28 Based on our results, adolescents with only primary education suffer more frequently than others from mental disorders. Almost a third of the cohort members who had not completed secondary level education suffered from mental disorders. As such, short education, mental disorders and financial difficulties often accumulate in early adulthood. In this study, age at the time of the first psychiatric diagnosis was not studied, so it cannot be determined whether the psychiatric problem was the reason for short education or vice versa.

    The magnitudes of the risks between different determinants varied by sex. These variations may be due to the different ways that boys and girls respond to different situations, perhaps leading to different diagnoses and different levels of care referrals. These issues need to be studied further.

    The fact that 13% of the young adults had had a mental disorder during the 21-year follow-up does not present the true prevalence of lifetime mental disorders, since we know that many children and adolescents with mental disorders remain unrecognised and untreated.29 ,30 Also, the outpatient care data are available from 1998, so those children diagnosed in outpatient psychiatric care before 1998 and with no later visits are missing from the data. Thus, the true prevalence of mental disorders would be even higher than that which we present. In this study, the mental disorder diagnoses are based on the use of specialised care, meaning that these disorders are severe; less difficult psychological symptoms are treated in primary healthcare in Finland. In addition, we do not have data on private outpatient child and adolescent psychiatric services. However, in Finland, most children use public services through child and school health services, which are provided by the municipalities for all children and adolescents free of charge.

    Our results tell about the influence of parents’ financial and psychiatric problems on the mental disorders of their children. Also, based on this and our previous research,16 parents’ low educational level and low socioeconomic position increase the risk for children's mental disorders. Public specialised psychiatric services were used significantly more by children and adolescents from lower socioeconomic and educational backgrounds. Since the mental disorders are based on the use of specialised care, those cohort members who have not received care are missing from the analysis. That may explain why low parental education and low SES show up as protective factors within the full model. That may also explain the finding that a cohort member's short education was a more significant determinant for mental disorder when the parents were educated than when both the parents and the cohort member had only a basic education. The other possibility is that this is due to different psychiatric morbidity. Different explanations need further studies. In our previous article,16 we showed that public specialised psychiatric services were used significantly more by children and adolescents from lower socioeconomic and educational backgrounds. In that study, however, familial or social factors were not assessed. Now, taken together, these studies suggest that children with parents who are more educated and have better socioeconomic position may have better access to care. This question cannot be answered since the register data do not provide information about the care seeking patterns for psychiatric care or on the total morbidity. However, equity in access to mental health services is essential and should be targeted in the health and social welfare policies.

    One of the main strengths of our study is that it involves a complete census of all infants born in a single year in Finland. The follow-up covers the time from birth to early adulthood and therefore also provides information about later mental disorders beyond childhood. It is well known that mental disorders often occur in late adolescence or early adulthood, even though their roots may go back to childhood.4 Another significant strength of our study is its data coverage, both in terms of study population and variables. Survey studies related to mental disorders often have difficulties in reaching exactly those whose welfare we know the least about, as willingness to take part in research is related to SES and, as a result, often those with low SES are under-represented in surveys. The data are based on Finnish national registers and thus not biased by background variables.

    While our results show significant intergeneration transfers of disadvantage, this phenomenon is not deterministic in nature. Children's and young people’s mental disorders may be partly explained by genetic and neurological factors that we cannot in retrospect alter, but it is possible to influence some of the environmental factors such as family relations, school performance, social relationships or hobbies and, as such, support children's well-being. Families with social adversities and with parental mental health problems should be supported to secure children's development and to prevent marginalisation. Instead of stepping in when problems have already appeared, it would be essential to invest in well-being and strengthening of resilience, efforts in which the role of day care and school get emphasised. In addition, investing in primary services as well as in outpatient psychiatric services is crucial. Preventive work and early interventions are crucial not only from a financial perspective, but also from a human perspective.

    What is already known on this subject?

    • Mental disorders are a significant concern for public health. Most mental disorders have roots in childhood and various risk factors for mental disorders have been identified. However, the life-course development of mental disorders and intergenerational transmission of inequalities in well-being is not sufficiently known.

    What this study adds?

    • This study with a longitudinal nationwide follow-up through childhood and adolescence shows that mental disorders have many social risk factors which are interlinked. In addition, low education, mental disorders and financial difficulties accumulate in early adulthood. There is a significant intergenerational transmission of disadvantage in a modern welfare society. This study provides evidence on the factors that threaten mental health. This knowledge is needed for comprehensive, preventative and supporting efforts.

    References

    1. ↵
      1. Kessler RC,
      2. Berglund P,
      3. Demler O,
      4. et al
      . Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 2005;62:593–602.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Patel V,
      2. Flisher AJ,
      3. Hetrick S,
      4. et al
      . Mental health of young people: a global public-health challenge. Lancet 2007;369:1302–13.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Wahlbeck K,
      2. Westman J,
      3. Nordentoft M,
      4. et al
      . Outcomes of Nordic mental health systems: life expectancy of patients with mental disorders. Br J Psychiatry 2011;199:453–8.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Fryers T,
      2. Brugha T
      . Childhood determinants of adult psychiatric disorder. Clin Pract Epidemiol Ment Health 2013;9:1–50.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Robinson M,
      2. Oddy WH,
      3. Li J,
      4. et al
      . Pre- and postnatal influences on preschool mental health: a large-scale cohort study. J Child Psychol Psychiatry 2008;49:1118–28.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Thompson R,
      2. Lindsey MA,
      3. English DJ,
      4. et al
      . The influence of family environment on mental health need and service use among vulnerable children. Child Welfare 2007;86:57–74.
      OpenUrlPubMedWeb of Science
    7. ↵
      1. Patterson PH
      . Neuroscience. Maternal effects on schizophrenia risk. Science 2007;318:576–7.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Veijola J,
      2. Laara E,
      3. Joukamaa M,
      4. et al
      . Temporary parental separation at birth and substance use disorder in adulthood. A long-term follow-up of the Finnish Christmas Seal Home Children. Soc Psychiatry Psychiatr Epidemiol 2008;43:11–17.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Weaver IC
      . Shaping adult phenotypes through early life environments. Birth Defects Res C Embryo Today 2009;87:314–26.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Ekblad M,
      2. Gissler M,
      3. Lehtonen L,
      4. et al
      . Prenatal smoking exposure and the risk of psychiatric morbidity into young adulthood. Arch Gen Psychiatry 2010;67:841–9.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Lahti M,
      2. Raikkonen K,
      3. Wahlbeck K,
      4. et al
      . Prenatal origins of hospitalization for personality disorders: the Helsinki birth cohort study. Psychiatry Res 2010;179:226–30.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Raikkonen K,
      2. Lahti M,
      3. Heinonen K,
      4. et al
      . Risk of severe mental disorders in adults separated temporarily from their parents in childhood: the Helsinki birth cohort study. J Psychiatr Res 2011;45:332–8.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Gissler M,
      2. Haukka J
      . Finnish health and social welfare registers in epidemiological research. Norsk Epidemiol 2004;14:113–20.
      OpenUrl
    14. ↵
      1. Paananen R,
      2. Gissler M
      . Cohort profile: the 1987 Finnish Birth Cohort. Int J Epidemiol 2012;41:941–5.
      OpenUrlFREE Full Text
    15. ↵
      1. Keskimaki I,
      2. Aro S
      . Accuracy of data on diagnosis, procedures and accidents in the Finnish Hospital Discharge Register. Int J Health Sci 1992;2:15–21.
      OpenUrl
    16. ↵
      1. Paananen R,
      2. Santalahti P,
      3. Merikukka M,
      4. et al
      . Socioeconomic and regional aspects in the use of specialized psychiatric care—a Finnish nationwide follow-up study. Eur J Public Health 2013;23:372–7.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Duncan GJ,
      2. Brooks-Gunn J
      . Consequences of growing up poor. New York: Russell Sage Foundation, 1998.
    18. ↵
      1. Pearce A,
      2. Lewis H,
      3. Law C
      . The role of poverty in explaining health variations in 7-year-old children from different family structures: findings from the UK Millennium Cohort study. J Epidemiol Community Health 2013;67:181–9.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Beardslee WR,
      2. Versage EM,
      3. Gladstone TR
      . Children of affectively ill parents: a review of the past 10 years. J Am Acad Child Adolesc Psychiatry 1998;37:1134–41.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Vandeleur C,
      2. Rothen S,
      3. Gholam-Rezaee M,
      4. et al
      . Mental disorders in offspring of parents with bipolar and major depressive disorders. Bipolar Disord 2012;14:641–53.
      OpenUrlCrossRef
      1. Weissman MM,
      2. Wickramaratne P,
      3. Nomura Y,
      4. et al
      . Offspring of depressed parents: 20 years later. Am J Psychiatry 2006;163:1001–8.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Ashford J,
      2. van Lier PA,
      3. Timmermans M,
      4. et al
      . Prenatal smoking and internalizing and externalizing problems in children studied from childhood to late adolescence. J Am Acad Child Adolesc Psychiatry 2008;47:779–87.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Stene-Larsen K,
      2. Borge AI,
      3. Vollrath ME
      . Maternal smoking in pregnancy and externalizing behavior in 18-month-old children: results from a population-based prospective study. J Am Acad Child Adolesc Psychiatry 2009;48:283–9.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Cornelius MD,
      2. Goldschmidt L,
      3. De Genna NM,
      4. et al
      . Long-term effects of prenatal cigarette smoke exposure on behavior dysregulation among 14-year old offspring of teenage mothers. Matern Child Health J 2012;16:694–705.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Müller KU,
      2. Mennigen E,
      3. Ripke S,
      4. et al
      . Altered reward processing in adolescents with prenatal exposure to maternal cigarette smoking. JAMA Psychiatry 2013;19:1–10.
      OpenUrl
      1. Moylan S,
      2. Jacka FN,
      3. Pasco JA,
      4. et al
      .  How cigarette smoking may increase the risk of anxiety symptoms and anxiety disorders: a critical review of biological pathways. Brain  Behav 2013;3:302–26.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Kessler RC
      . The costs of depression. Psychiatr Clin North Am 2012;35:1–14.
      OpenUrlCrossRefPubMed
    28. ↵
      1. Breslau J,
      2. Lane M,
      3. Sampson N,
      4. et al
      .  Mental disorders and subsequent educational attainment in a US national sample. J Psychiatr Res 2008;42:708–16.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Zwaanswijk M,
      2. Verhaak PF,
      3. Bensing JM,
      4. et al
      . Help seeking for emotional and behavioural problems in children and adolescents: a review of recent literature. Eur Child Adolesc Psychiatry 2003;12:153–61.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Suvisaari J,
      2. Aalto-Setälä T,
      3. Tuulio-Henriksson A,
      4. et al
      . Mental disorders in young adulthood. Psychol Med 2009;39:287–99.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • Contributors PR, RT and GM were involved in the conception, design and interpretation of the data. MM was involved in the analyses. All authors contributed to the drafting of the article and to intellectual content.

    • Funding The Academy of Finland, Emil Aaltonen Foundation and National Institute for Health and Welfare, Finland.

    • Competing interests None.

    • Ethics approval Finnish National Institute for Health and Welfare Ethical committee.

    • Provenance and peer review Not commissioned; externally peer reviewed.